AMBER Archive (2009)

Subject: [AMBER] ongoing issues using sander with AMOEBA

From: Tomasio, Susana (S.M.T.Tomasio_at_warwick.ac.uk)
Date: Wed May 13 2009 - 05:00:00 CDT

Hi Wei,

  We are still experiencing the problem outlined below when we try to use sander with AMOEBA. We followed your advice given below (regarding running of the test jobs relating to Run.Amoeba_sol). To re-iterate:

the job starts to run, but mdinfo remains an empty file. The hpv.out file is not written to beyond the point where it prints out
 TOTAL SIZE OF NONBOND LIST = 4359315

i.e. just before commencing the actual md part of the job, nothing else is written to file.

We note this regardless of the run time we allocate (e.g. even after 8 hours this situation has not progressed). Further, we tried allocating different amounts of memory - but both 1 Gb and 4Gb allocations yielded the same result.

Should this particular job really take a lot of time? Is it reasonable to expect the file to be written to within 8 hours? The hardware we are running this on comprises 960 x 3 GHz Intel Xeon "Woodcrest" cores arranged as 240 x 2-way dual core nodes.

Is there another flag we can add to the command line to get a more verbose output in order to track down our problem?

We would really appreciate help with getting this issue resolved.

thanks

Susana Tomasio and Dr. Tiff Walsh

-----Original Message-----
From: amber-bounces_at_ambermd.org on behalf of Tomasio, Susana
Sent: Wed 4/29/2009 9:15 PM
To: AMBER Mailing List
Subject: RE: [AMBER] test/amoeba: forrtl: error (78): process killed (SIGTERM)
 
Hi Wei

Thank you for your reply.

I tried your suggestion but I donīt think itīs running properly. In the qstat it says that it is running,
however the mdinfo output is empty and the results section of the file hpv.out is the following: (it doesnīt change
with time)

| # of SOLUTE degrees of freedom (RNDFP): 42414.
| # of SOLVENT degrees of freedom (RNDFS): 0.
| NDFMIN = 42411. NUM_NOSHAKE = 0 CORRECTED RNDFP = 42411.
| TOTAL # of degrees of freedom (RNDF) = 42411.
 ---------------------------------------------------
 APPROXIMATING switch and d/dx switch using CUBIC SPLINE INTERPOLATION
 using 5000.0 points per unit in tabled values
 TESTING RELATIVE ERROR over r ranging from 0.0 to cutoff
| CHECK switch(x): max rel err = 0.2738E-14 at 2.422500
| CHECK d/dx switch(x): max rel err = 0.8987E-11 at 2.875760
 ---------------------------------------------------
| Local SIZE OF NONBOND LIST = 2500071
| TOTAL SIZE OF NONBOND LIST = 4359315

Thank you,

Susana

-----Original Message-----
From: amber-bounces_at_ambermd.org on behalf of Wei Zhang
Sent: Wed 4/29/2009 7:00 PM
To: AMBER Mailing List
Subject: Re: [AMBER] test/amoeba: forrtl: error (78): process killed (SIGTERM)
 
Hi Susana,

    There are two test cases under that directory: Run.amoeba and
Run.amoeba_sol. Can you
try to run sander using the output of the other one: Run.amoeba_sol?
The command should
be:

     $AMBERHOME/exe/sander -O -i mdin -c hpv.xyz -p hpv.prmtop.save -o
hpv.out

     Sincerely,

     Wei

On Apr 29, 2009, at 12:45 PM, Tomasio, Susana wrote:

>
> Hi all!!
>
> I'm still stuck with the same problem! I am just trying to run the
> amoeba test and I can't run sander.
> I don't understand where does the file 'inpcrd' that is on
> $AMBERHOME/amber10/test/sleap/amoeba come from!
> After passing the Run.amoeba test this is the error that I get when
> I try to run sander:
>
> The command that I am using is the following:
>
> $AMBERHOME/exe/sander.MPI -O -i mdin -o jac.out -c jac.xyz -p
> jac_2.prmtop -r jac.rst -x jac.mdcr
>
> bash: mc: line 1: syntax error: unexpected end of file
> -bash: error importing function definition for `mc'
> [node11:18813] MPI_ABORT invoked on rank 0 in communicator
> MPI_COMM_WORLD with errorcode 1
> bash: mc: line 1: syntax error: unexpected end of file
> -bash: error importing function definition for `mc'
> [node11:18813] MPI_ABORT invoked on rank 0 in communicator
> MPI_COMM_WORLD with errorcode 1
> forrtl: error (78): process killed (SIGTERM)
> Image PC Routine Line
> Source
> libpthread.so.0 00002AD2F7C53C00 Unknown Unknown
> Unknown
> libopen-pal.so.0 00002AD2F720194C Unknown Unknown
> Unknown
> libmpi.so.0 00002AD2F6D99784 Unknown Unknown
> Unknown
> libmpi.so.0 00002AD2F6DF9ED5 Unknown Unknown
> Unknown
> libmpi.so.0 00002AD2F6DFAABB Unknown Unknown
> Unknown
> libmpi.so.0 00002AD2F6DF46CA Unknown Unknown
> Unknown
> libmpi.so.0 00002AD2F6DAE4AA Unknown Unknown
> Unknown
> libmpi_f77.so.0 00002AD2F6C31BC4 Unknown Unknown
> Unknown
> sander.MPI 00000000004F1863 Unknown Unknown
> Unknown
> sander.MPI 00000000004EEAC6 Unknown Unknown
> Unknown
> sander.MPI 000000000043FE22 Unknown Unknown
> Unknown
> libc.so.6 00002AD2F7D7C184 Unknown Unknown
> Unknown
> sander.MPI 000000000043FD69 Unknown Unknown
> Unknown
>
> Image PC Routine Line
> Source
> libpthread.so.0 00002AD2F7C53C00 Unknown Unknown
> Unknown
> libopen-pal.so.0 00002AD2F720194C Unknown Unknown
> Unknown
> libmpi.so.0 00002AD2F6D99784 Unknown Unknown
> Unknown
> libmpi.so.0 00002AD2F6DF9ED5 Unknown Unknown
> Unknown
> libmpi.so.0 00002AD2F6DFAABB Unknown Unknown
> Unknown
> libmpi.so.0 00002AD2F6DF46CA Unknown Unknown
> Unknown
> libmpi.so.0 00002AD2F6DAE4AA Unknown Unknown
> Unknown
> libmpi_f77.so.0 00002AD2F6C31BC4 Unknown Unknown
> Unknown
> sander.MPI 00000000004F1863 Unknown Unknown
> Unknown
> sander.MPI 00000000004EEAC6 Unknown Unknown
> Unknown
> sander.MPI 000000000043FE22 Unknown Unknown
> Unknown
> libc.so.6 00002AD2F7D7C184 Unknown Unknown
> Unknown
> sander.MPI 000000000043FD69 Unknown Unknown
> Unknown
>
> Does anyone can help me please?
>
> Thank you,
>
> Susana
> _______________________________________________
> AMBER mailing list
> AMBER_at_ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber

_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber

_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber

_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber